MATLAB Implementation of Convolutional Code Encoding and Decoding with Soft Decision Decoding

This MATLAB program demonstrates the complete process of convolutional code encoding and decoding. For encoding implementation, we generate random binary sequences as source input and utilize MATLAB's poly2trellis function to create the trellis structure defining the convolutional code parameters. The convenc function is then employed to perform the actual encoding operation, which involves shift registers and generator polynomials to produce coded output.

For the decoding phase, we implement soft decision Viterbi decoding using the vitdec function with appropriate parameters for continuous operation mode. This approach processes quantized analog voltage values from the channel instead of hard-decoded bits, providing 3-4 dB performance improvement over hard decision decoding. The algorithm maintains path metrics and uses trellis traceback to determine the most likely transmitted sequence.

Through this implementation, users can gain practical understanding of convolutional code operational principles, including code rate constraints, memory depth effects, and trellis path management. Performance analysis can be conducted by modifying key parameters such as constraint length, generator polynomials, and quantization levels for soft decisions to optimize bit error rate performance under various channel conditions.